2.0 Analysis 2.1 Introduction No mechanical discrepancies were found with the aircraft that would have contributed to the occurrence; the analysis will, therefore, focus on the combination of factors that led to the runway excursion. 2.2 The Approach and Landing Had weather conditions not required the crew to manoeuvre around low cloud on short final, it is probable that the aircraft would have been on a stabilized approach. This would have enabled the co-pilot to land before 1,650 feet from the threshold, and he might have been able to stop the aircraft on the available runway. The tailwind component present during the approach and landing contributed to the extended landing distance and additional ground run of the aircraft. 2.3 Crew Decisions On breaking out of the cloud, the crew saw that a straight-in landing was not possible. They had briefed for this possibility, and elected to conduct a circling approach with a left turn to the other end of the runway. Several factors tended to de-stabilize the approach. The visual circling procedure had to be conducted with limited visibility in a confined area in freezing precipitation. The crew had to be mindful of the high terrain. There was a tailwind component, and the aircraft had to be manoeuvred around low cloud close to the runway threshold. The BAe 146 depends on brakes and lift dumping devices for deceleration. These may be less effective than reverse thrust in the high-speed early parts of a landing roll on a slippery surface. The captain was faced with letting the co-pilot develop his judgement in challenging circumstances or controlling the aircraft himself. When the co-pilot lost sight of the runway because his view of it was blocked by his cross-cockpit line of sight, the captain would have been justified in taking control of the aircraft. Three times on the final approach, the captain suggested that they go around, but he left the decision with the co-pilot. The co-pilot expressed doubts about being able to see the runway again if they attempted another approach and continued with the approach and landing. After the crew landed, they encountered unexpected conditions. The runway was more slippery than they had expected. In part, they relied too much on the JBI information provided to them and the pilot report from the MU-2. The JBI readings were taken on a track outside the wheel path of the aircraft on what was likely a less slippery surface. The MU-2 had a narrow track and might also have encountered different conditions than the landing BAe-146. In part, the published JBI information does not make clear that the normal landing distances safety factor is not present for the lower JBI readings. Neither does it express that reduced braking efficiency at lower JBIvalues makes the figures non-conservative. There are cautionary words to make clear that JBIvalues are only for guidance. The BAe manufacturer's operating manual, if consulted, would have shown that remaining on the runway would have been problematic. The BAe information was made available to Air BC in 1993, and became a flight manual amendment. In a situation where there is very little room for departure from the ideal landing profile, the effects of several factors increased the aircraft's landing roll to the extent that the aircraft overran the end of the runway. The crew could have and should have been aware of some of the effects, but they had no way of knowing about others, including the up to 18 per cent error in JBI readings and the reduction in the landing distance safety factor. The aircraft landed about 1,650 feet beyond the runway threshold, which is about 650 feet beyond the ideal touchdown point. 2.4 JBI Calculations Because urea is spread on the runway centre area first, the centre of the runway might have been wetter than the surface 30 feet either side of the centre line, where the JBI was measured. The centre of the runway, where the BAe 146 landed, probably had a lower JBI than that reported. This would have caused the ineffective braking reported by the co-pilot. Further, as JBI figures are only accurate for packed snow or ice, and not for wet or slush-covered runways, it was inappropriate for the pilots to rely on the JBI table to calculate the aircraft's stopping distance. The JBI information published by TC in the AIP is for guidance only. The information is misleading as it does not give sufficient warning that the runway requirement figures for low JBIconditions have not been validated and contain inaccuracies of up to 18 per cent. Although transport category aircraft such as the BAe 146 are given type approval which specifies that the aircraft shall be able to stop within 60 per cent of the intended landing field length, the use of the JBI table enables an operator to legally land on a contaminated runway where the JBI table predicts that the aircraft will require the full runway length, with no safety margin. Hence the 40per cent safety margin is lost when it is most needed--that is, when the runway is contaminated and the stopping distance is unpredictable. Some pilots are not aware of this limitation regarding the JBI table and believe that the 60 per cent safety factor is still present. 2.5 BAe Contaminated Landing Performance Figures Compared to JBIFigures The incident pilots relied unduly on the JBI figures, and did not realize that the safety margin of the 60 per cent factor had been removed, that the JBI figures were up to 18percent inaccurate, and that the JBI figures were estimates, to be used only as a guide. Landing distance calculations using the figures in section 9.10.11 of the BAe 146 Operations Manual (MOM) indicate that the aircraft would run off the end of the runway under the prevailing conditions; however, Air BC does not train, or require, its pilots to use the MOM performance charts to determine landing distances. Landing distances predicted by both the MOM charts and the Air BC charts are longer than those predicted by the JBI tables; it was not possible, however, to determine the reason for this discrepancy, because the engineering data upon which the JBI table was based are not available. 3.0 Conclusions 3.1 Findings The crew had to manoeuvre the aircraft around low cloud on short final to maintain visual contact with the runway. The aircraft landed about 1,650 feet beyond the runway threshold, which is about 650feet beyond the ideal touchdown point. The aircraft landed with a tailwind component on a downsloped runway. The aircraft went off the end of the runway at a speed of approximately 25 knots. There was low cloud, freezing precipitation, and a tailwind when the aircraft landed. The captain suggested to the co-pilot that they go around, but the co-pilot expressed doubts about being able to see the runway again if they attempted another approach and continued with the approach and landing. The 60 per cent factor is removed from the JBI table, and this is not clearly stated in theAIP. The JBI readings sample only portions of the runway. TC has no record to validate the JBI figures, and a recent TC study found that actual landing distances were as much as 18 per cent greater than the AIP values. The crew relied unduly on the JBI table. There was a discrepancy between the BAe 146 landing performance figures of BAe and those of Air BC. The aircraft was serviced and maintained in accordance with existing directives. Its weight and balance were within limits and the centre of gravity was within the normal range. The pilots were certificated and qualified for the flight in accordance with existing regulations. The BAe 146 aircraft is not equipped with thrust reversers and relies on brakes and lift dumping devices which are less effective on slippery surfaces. 3.2 Causes The crew flew a landing pattern in a narrow valley under marginal weather conditions that dictated an unstabilized approach to the slippery and contaminated runway. The result was a longer-than-normal landing and landing roll. Contributing to the occurrence were a tailwind component, inaccurate James Brake Index (JBI) tables, crew decision making, and the company's use of landing performance tables different from those produced by the aircraft manufacturer for contaminated runways. 4.0 Safety Action 4.1 Action Taken 4.1.1 JBI Tables In light of this occurrence, the TSB sent an Aviation Safety Advisory (950056) to Transport Canada (TC) in April 1995. The Advisory highlighted the need for TC to re-evaluate the JBItables and revise the information portrayed in the AIP accordingly. Additionally, it stated that TC might wish to consider interim action to inform the aviation community of the apparent shortcomings of the JBI tables. Subsequent to the TSB Advisory, Air BC indicated other concerns regarding the use of landing distances based on JBI, namely the appropriateness of the runway locations where JBI measurements are taken, the need for JBI corrections for all possible runway conditions, and the possible adverse effect on stopping distances of melting due to urea application. The TSB forwarded these concerns to TC. In addition, a second Aviation Safety Advisory (960084) was sent to TC in August 1996 on the lack of an explicit indication that the JBItables do not include the 60 per cent safety factor in the derived landing distances for contaminated runways. TC has indicated that no changes will be made to the JBI tables until the results of a combined NASA/FAA/TC winter runway test program are available. TC also indicated that an Air Carrier Advisory Circular will be released on the issue of contaminated runways. 4.1.2 Operator Action Air BC has indicated that, until the JBI data is validated, 20 per cent will be added to any required landing distance when using a JBI correction factor from the applicable table in the AIP.